Stats Lie

[The Big Cat]

Is there a comprehensive statistic that takes game-situations into account? If there isn't, there should be.

 

At the end of the game, the Bills were glad to give up 5-10 yard chunks in order to leverage the game clock against the Dolphins. Either way, Tannehill's final numbers look rather mediocre, but anybody who watched the game knows that we shut him down.

 

I'm sure football outsiders have developed some newfangled statistical system, but I agree with you, and have given it much thought, I'm just not a stats guy.

 

Seems like you should be able to create a model that weighs time remaining, point margin, down, down distance and others to compute the stats that matter.

[K-9]

I know the threshold isn't down 35-3.

 

Remember the Ravens/Vikings game last year? Didn't they put up ~28 points combined in 3 minutes?

 

It's also important to remember that garbage time stats help the defenses sometimes too. Sacks and picks on hail mary's happen all the time. Percentage of victory shouldn't disregard those numbers.

 

 

 

But that's my point, it's only inconsequential after the fact. Fitz and the defense padded their stats down 21-0 vs the Pats in 2011. Until the end of the game, we had no idea if they were "garbage time" or not. At what point do you consider the outcome "decided?" What percentage of possible victory? 1%? Less than 1%? The possibility of victory will almost never be 0% until the clock has expired.

 

Of course it's after the fact. If Miami had busted a 70 yard TD run at the end of the 1st half or somehow managed to score and recover a couple onside kicks and score again late in the 4th quarter, then those stats aren't rendered meaningless and obviously wouldn't be categorized as such. Yes weird schit happens. But that's not what we're arguing, is it? I thought we were talking about padded stats that had absolutely NO impact on the outcome of the game.

 

GO BILLS!!!

[FireChan]

Of course it's after the fact. If Miami had busted a 70 yard TD run at the end of the 1st half or somehow managed to score and recover a couple onside kicks and score again late in the 4th quarter, then those stats aren't rendered meaningless and obviously wouldn't be categorized as such. Yes weird schit happens. But that's not what we're arguing, is it? I thought we were talking about padded stats that had absolutely NO impact on the outcome of the game.

 

GO BILLS!!!

 

I get what you're saying, I do. But if you're going to "throw out" stats as having no impact, you need a threshold of their possibility of impact. The Wallace TD also didn't change the outcome of the game, but we aren't throwing it out because the game wasn't for sure won. So when is the game for sure won? Let's say Miami had a 30% chance of winning after the Wallace TD. And then we got a field goal, lowering it to 15%. Is everything after 15% a non-impact stat? What about after the second field goal, lowering it to 5% Is everything after 5% a non-impact stat? These are just random percentages, but there would need to be a system and a threshold if we were going to throw away every team's "garbage stats" and then compare numbers.

Edited September 15, 2014 by FireChan

[K-9]

Is there a comprehensive statistic that takes game-situations into account? If there isn't, there should be.

 

At the end of the game, the Bills were glad to give up 5-10 yard chunks in order to leverage the game clock against the Dolphins. Either way, Tannehill's final numbers look rather mediocre, but anybody who watched the game knows that we shut him down.

 

^This. All those completions by Tannehill underneath were as good as rushing attempts vs. our prevent shell. It's ECACTLY what you want to see as a DC. I was loving it. If teams want to run the clock out on themselves, more power to them.

 

But doing that at the end of the 1st half was an admission by Philbin that his offense was getting owned. On second thought, it wasn't so much as an admission as it was a submission.

 

GO BILLS!!!

[Hapless Bills Fan]

The thing about garbage stats is that most of them are only garbage in hindsight. If Miller breaks a long run at the end of the half, and the Phins score, those runs are no longer garbage.

 

You play 60 minutes, and every play matters. There aren't really "garbage" stats.

 

This.

 

I would like someone to provide me with a proactive definition of "garbage time" and "garbage stats".

 

Can't be done, because like he said - it's retrospective. Someone out in San Fran was surely smug about the "garbage time" plays Chicago was running - until the Bears snuck ahead. Take your foot off the gas and get smug about "garbage time" and the team just may pay.

[The Dean]

Torture numbers long enough, and they'll admit to anything.

 

In truth, some stats are a bit more predictive than others. And stats can only be evaluated in context. Stats aren't useless. But empty stats don't reveal all that much.

Edited September 15, 2014 by The Dean

[Hapless Bills Fan]

I get what you're saying, I do. But if you're going to "throw out" stats as having no impact, you need a threshold of their possibility of impact. The Wallace TD also didn't change the outcome of the game, but we aren't throwing it out because the game wasn't for sure won. So when is the game for sure won?

 

I have an idea. Let's ask the Houston Oilers, ha ha!

[dave mcbride]

What kills me is the offseason appeal to authority that comes in the form of digging up stats. Both sides do it. Yet, I think we're all aware that football, unlike baseball, for example, does not provide sample sets that are nearly large enough to cover the qualitative importance that so many of the numbers carry.

 

Over a 16 game season, that stuff usually evens out. The teams with the most yards definitely tend to score the most points, and the teams that give up the least yards tend to give up the lowest number of points. Just look at the Steelers' defensive rankings for points and yards over the past 20 years - they have a close correlation in almost every year. http://www.pro-football-reference.com/teams/pit/

[K-9]

I get what you're saying, I do. But if you're going to "throw out" stats as having no impact, you need a threshold of their possibility of impact. The Wallace TD also didn't change the outcome of the game, but we aren't throwing it out because the game wasn't for sure won. So when is the game for sure won?

 

I don't agree that I need a "threshold of their possibility of impact" because it's all ascertained after the fact. A game is won "for sure" when you have more points than the other team when the clock reads "00." But if you are a purist that insists there's no such thing as "garbage time", more power to you.

 

GO BILLS!!!

[Max997]

Great example of why looking at yards and other stats in a vacuum tells you virtually nothing.

 

At the end of the first half, Miami ran six pointless carries for 29 pointless yards. Aside from getting one of our defenders injured, I truly fail to see the point in any of these plays.

 

Stricken from the final stat sheet, and our boys surrendered only 51 yards on 15 carries.

 

Likewise, down 19 points at the end of the game, the Dolphins ran 9 more pointless plays, racking up another 60, pointless yards. Now, granted, wiping those numbers from the sheet also negate an interception, but regardless, if not for those garbage plays, Tannehill (whom every last pundit has signed off on being forever superior to EJ) went 25/40 for 180 yards. That's just over 7 ypc, and just over 4 ypa.

 

I say this now, because come May, there will be arguments aplenty, and both sides will be waving 2014 stat results around like they actually mean something.

 

I wish we could compile a stat book for these arguments that just took garbage stats and burned them.

 

couldnt agree more

 

basing an opinion on anything other then what you see with your own eyes is a completely ridiculous and flawed opinion

[Over 29 years of fanhood]

Great example of why looking at yards and other stats in a vacuum tells you virtually nothing.

 

At the end of the first half, Miami ran six pointless carries for 29 pointless yards. Aside from getting one of our defenders injured, I truly fail to see the point in any of these plays.

 

Stricken from the final stat sheet, and our boys surrendered only 51 yards on 15 carries.

 

Likewise, down 19 points at the end of the game, the Dolphins ran 9 more pointless plays, racking up another 60, pointless yards. Now, granted, wiping those numbers from the sheet also negate an interception, but regardless, if not for those garbage plays, Tannehill (whom every last pundit has signed off on being forever superior to EJ) went 25/40 for 180 yards. That's just over 7 ypc, and just over 4 ypa.

 

I say this now, because come May, there will be arguments aplenty, and both sides will be waving 2014 stat results around like they actually mean something.

 

I wish we could compile a stat book for these arguments that just took garbage stats and burned them.

 

There is the one thing that supersedes and makes stats somewhat less relevant. It's a scoreboard, and it's been doing a pretty good job of resolving any ambiguity in the stats which ultimately don't lie at all as long as one understands what the stats are actually telling them.

[ExiledInIllinois]

What I noticed... Time of possesion and first downs. BFLO in both games got beat in those columns... That does have to change.

[FireChan]

I don't agree that I need a "threshold of their possibility of impact" because it's all ascertained after the fact. A game is won "for sure" when you have more points than the other team when the clock reads "00." But if you are a purist that insists there's no such thing as "garbage time", more power to you.

 

GO BILLS!!!

 

Exaclty. I suppose the disconnect lies in TBC saying "For a while I've thought about what would constitute "garbage" time, and it seems there would have to be a ratio of point margin to seconds left in the game. There is a threshold from which no team (yet) has come back. I'm very curious to know what it is."

 

That's more what I was saying, and you would need a ratio or system to eliminate the "garbage time" stats when comparing two team's stats. And as no system really exists, there's less room for error when you just use all the stats and assume garbage time is a non-factor.

[Miyagi-Do Karate]

Big cat: great post.

 

I don't know why we have to extrapolate this game to debate the contours of garbage time stats. I think the point here is in this particular game, the dolphins got a bunch of inconsequential yards that will, looking purely at stats, understate the bills dominance.

[FireChan]

Big cat: great post.

 

I don't know why we have to extrapolate this game to debate the contours of garbage time stats. I think the point here is in this particular game, the dolphins got a bunch of inconsequential yards that will, looking purely at stats, understate the bills dominance.

 

His point was clear, it just organically transformed. I can turn it into an EJ thread, if you'd rather read that.

[The Dean]

couldnt agree more

 

basing an opinion on anything other then what you see with your own eyes is a completely ridiculous and flawed opinion

 

 

Well in the words of Little Feat:

 

 

Hear me well

Seein' ain't always believin'

Just make sure it's the truth that you're seein'

Eyes sometimes lie, eyes sometimes lie

They can be real deceivin'

 

 

Truth is, you, me, any one person, can only see so much of the field. We get confused as to how open a guy looked, how good/bad the throw was. We also have out own particular preconceptions and biases. Of course just using stats, and not seeing for yourself is crazy. But, IMO, only using your eyes to judge is nearly as flawed a method of analysis.

[Dorkington]

What I noticed... Time of possesion and first downs. BFLO in both games got beat in those columns... That does have to change.

 

Turnovers and special teams plays have played a part in us being behind in both of those columns, to be fair.

 

For example, we didn't have the ball until late in the 3rd quarter because of CJ's TD return.

 

Also, we've had the shortest field on average to start with:

 

http://www.pro-football-reference.com/years/2014/ (scroll down to Drives)

Edited September 15, 2014 by Dorkington

[NoSaint]

 

 

I'm sure football outsiders have developed some newfangled statistical system, but I agree with you, and have given it much thought, I'm just not a stats guy.

 

Seems like you should be able to create a model that weighs time remaining, point margin, down, down distance and others to compute the stats that matter.

 

 

espn toyed with it for their total QBR

[billsfan89]

The thing about garbage stats is that most of them are only garbage in hindsight. If Miller breaks a long run at the end of the half, and the Phins score, those runs are no longer garbage.

 

You play 60 minutes, and every play matters. There aren't really "garbage" stats.

 

Garbage stats don't really exist in the first 3 quarters. Any play could result in a big play that helps to turn the momentum in a game. I think if one team has a big lead the last offensive drive or two (Depending on the size of the lead) is garbage time. Outside of those last couple drives in a blowout garbage time is overblown.

[enlightener]

all games have these same garbage stats rolled into them, or some will have a lot while some have almost every play count. in the end the stats till tell you a lot depending on what you are looking for and how you read them, like if the bills were good offensive team last year passing vs running or our pass defense vs run defense, or if SJ was ever a deep threat or was Evans a better all around WR than moulds etc....